A model of autosomal recessive Alport syndrome in English cocker spaniel dogs

Kidney International, Vol. 54 (1998), pp. 706 –719

A model of autosomal recessive Alport syndrome in English cocker spaniel dogs GEORGE E. LEES, R. GAYMAN HELMAN, CLIFFORD E. KASHTAN, ALFRED F. MICHAEL, LINDA D. HOMCO, NICHOLAS J. MILLICHAMP, YOSHIFUMI NINOMIYA, YOSHIKAZU SADO, ICHIRO NAITO, and YOUNGKI KIM Texas Veterinary Medical Center, Texas A&M University, College Station, Texas, Oklahoma Animal Disease Diagnostic Laboratory, Oklahoma State University, Stillwater, Oklahoma, and Department of Pediatrics, University of Minnesota Medical School, Minneapolis, Minnesota, USA; Department of Molecular Biology and Biochemistry, Okayama University Medical School, Okayama, and Divisions of Immunology and Ultrastructural Biology, Shigei Medical Research Institute, Okayama, Japan

A model of autosomal recessive Alport syndrome in English cocker spaniel dogs. Background. Dogs with naturally occurring genetic disorders of basement membrane (type IV) collagen may serve as animal models of Alport syndrome. Methods. An autosomal recessive form of progressive hereditary nephritis (HN) was studied in 10 affected, 3 obligate carrier, and 4 unaffected English cocker spaniel (ECS) dogs. Clinical, pathological, and ultrastructural features of the disease were characterized. Expression of basement membrane (BM) proteins was examined with an immunohistochemical technique using monospecific antibodies. Results. Affected dogs had proteinuria and juvenile-onset chronic renal failure. Glomerular basement membrane (GBM) thickening and multilamellation typical of HN were observed in all renal specimens obtained from proteinuric dogs, and severity of GBM ultrastructural abnormalities varied with the clinical stage of disease. Expression of a3(IV) and a4(IV) chains was totally absent in the kidney of affected dogs. Expression of a5(IV) and a6(IV) chains was normal in Bowman’s capsule, collecting tubular BM and epidermal BM of affected dogs. The a5(IV) chain was not expressed in distal tubular BM of affected dogs. Expression of a5(IV) chains was markedly reduced but not absent, and expression of a6(IV) chains was present in GBM of affected dogs. Expression of a1-a2(IV) chains in GBM of affected dogs was increased. Features of obligate carriers were similar to those of unaffected dogs. Conclusions. We conclude that HN in ECS dogs is a naturally occurring animal model of autosomal recessive Alport syndrome. However, it differs from human disease in the persistence of a5(IV) chains in GBM and in the appearance of a6(IV) chains in GBM.

Hereditary nephritis (HN) refers to a group of genetic disorders of basement membrane (type IV) collagen that Key words: genetic disorder, chronic renal failure, hereditary nephritis, basement membrane, type IV collagen, progressive glomerular disease, multilamination. Received for publication August 11, 1997 and in revised form April 6, 1998 Accepted for publication April 14, 1998

© 1998 by the International Society of Nephrology

cause progressive glomerular disease [1–3]. In humans with Alport syndrome (AS), the nephropathy is frequently associated with sensorineural hearing loss and ocular abnormalities. Distinctive ultrastructural changes in glomerular basement membranes (GBM) of affected individuals is a prominent characteristic of these disorders [1–3]. In humans AS usually is X-linked, resulting from mutations in the COL4A5 gene, which encodes the a5 chain of type IV collagen [1, 4]. Some families have an autosomal recessive form of AS, due to mutations in the COL4A3 or COL4A4 gene on chromosome 2, that affect the a3(IV) or a4(IV) chain [5– 8]. Rare families have AS with an autosomal dominant pattern of inheritance [9]. Jefferson et al have mapped autosomal dominant AS to the region of COL4A3 and COL4A4 in a single family, but have yet to identify a specific mutation [10]. Naturally occurring diseases that are animal models for AS have been identified in dogs. An X-linked form of HN in a family of Samoyed dogs has been thoroughly characterized [11–16]. The causative mutation, located in exon 35 of the COL4A5 gene, is a single nucleotide substitution that produces a premature stop codon [17]. Additionally, autosomal dominant HN has been described in bull terrier dogs, but the causative mutation is unknown [18]. We have identified an autosomal recessive form of HN in English cocker spaniel (ECS) dogs. Initially, ultrastructural GBM lesions similar to those of HN were identified in several juvenile ECS dogs that had advanced renal failure [19]. Subsequently, repeated matings of the parents of affected dogs produced additional affected dogs [20]. Occurrence of an inherited nephropathy in ECS dogs has been recognized for many years; however, previous investigators had not clearly established that the disease was a form of HN [21–25]. Nonetheless, studies had shown that the disease was a rapidly progressive glomerular disorder that was inherited in an autosomal recessive fashion [26, 27]. Pedigrees of affected ECS dogs that we studied also were

706

707

Lees et al: Autosomal recessive Alport syndrome in dogs Table 1. English cocker spaniel dogs studied

Dog no.

Family

HN phenotype

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

A B C D E F G A H H A A I H J K L

Affected Affected Affected Affected Affected Affected Affected Affected Affected Affected Carrier Carrier Carrier Unaffected Normal Normal Normal

Age at initial exam Gender M M M F M M F M F F M F F M M F F

Age at last exam months

10 — 15 — 9 — — 3 3 3 — 59 102 3 — — —

Proteinuria 11 27 22 13 9 10 15 8 13 17 88 73 104 50 19 76 9

Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Neg Neg Neg Neg Neg Neg Neg

Renal failure

TEM thick GBM

Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Neg Neg Neg Neg Neg Neg Neg

Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos ND Neg Neg Neg Neg Neg Neg

Dogs 1 and 8 were siblings but not littermates; dogs 11 and 12 were parents of dogs 1 and 8. Dogs 9, 10, and 14 were littermate siblings. Abbreviations are: HN, hereditary nephritis; TEM, transmission electron microscopy; Affected, HN diagnosis based on clinicopathologic features and TEM findings; Carrier, parent of an HN-affected dog; Unaffected, healthy dog from HN-carrier parents; could be a carrier or a normal dog; Normal, kindred contains no HN-affected dogs, and little chance of being a HN-carrier; ND, not determined.

compatible with autosomal recessive transmission of the disease [20]. Immunohistochemical techniques have been used to study distribution of collagen a(IV) chains in basement membranes of humans with AS, dogs with HN, and transgenic mice with COL4A3 mutations. Expression of a3(IV), a4(IV), and a5(IV) usually is absent in the GBM of men with X-linked AS, while women who are heterozygous for X-linked AS mutations frequently lack these chains in portions of their GBM [28 –31]. In people with autosomal recessive AS, a3(IV), a4(IV), and a5(IV) chains usually are lacking in GBM; however, the a5(IV) chain is present in other basement membranes, for example, Bowman’s capsule, collecting tubular basement membrane, and epidermal basement membrane [32]. Transgenic COL4A3 mutant mice lack a3(IV), a4(IV) and a5(IV) chains in their GBM and tubular basement membranes [33, 34]. Human sera from patients with Goodpasture disease have been used to study X-linked HN in Samoyed dogs. These sera, which contain antibodies reacting mainly with a3(IV) chains, bind to normal dog GBM but not to the GBM of affected males [14]. Segmental GBM staining is seen in young heterozygous females, but staining becomes global as the dogs grow older [16]. In bull terrier dogs with autosomal dominant HN, renal expression of a3(IV) and a5(IV) chains appeared normal using monospecific antibodies [18]. This article describes the clinical, ultrastructural and immunohistochemical features of autosomal recessive HN in ECS dogs. Monospecific antibodies were used to deter-

mine the distribution of a(IV) chains in renal and epidermal basement membranes of affected dogs. METHODS Dogs Subjects were 17 ECS dogs that were selected in several ways (Table 1). Seven dogs (numbers 1 to 7) were identified when they developed juvenile-onset chronic renal failure. As described in a previous clinical report, which included four of these animals, such dogs were found be HNaffected by postmortem examinations that revealed characteristic ultrastructural GBM changes [19]. Four dogs (numbers 8 to 10, and 14) were offspring from repeat matings of dogs that had produced an affected dog in a previous litter. Three of these four dogs also were affected [20]. Three dogs (numbers 11 to 13) were obligate HN carriers, identified by having at least one offspring in which HN had been diagnosed. Three dogs (numbers 15 to 17) were healthy animals that served as controls. Each affected dog met two criteria for diagnosis of HN. First, the kidney disease exhibited clinicopathologic features identical to those described for the familial nephropathy that has been recognized in ECS dogs for many years. Second, examination with transmission electron microscopy (TEM) demonstrated characteristic abnormalities of GBM ultrastructure. Every dog with the clinicopathologic features of familial nephropathy that was examined with TEM had similar GBM changes.

708

Lees et al: Autosomal recessive Alport syndrome in dogs

Clinical evaluations

Table 2. Primary antibodies used in immunohistochemical studies

Eight dogs were evaluated once, and nine dogs were evaluated on 2 to 26 occasions. Five dogs with multiple evaluations were examined two to six times at irregular intervals. The other four dogs were evaluated monthly beginning when they were three months old and continuing until they developed renal failure and were euthanatized or became two years old and were still healthy [20]. Clinical evaluations were performed by conventional methods. Histories were reviewed, and complete physical examinations were performed. Renal ultrasonography was performed on 29 occasions in 15 dogs using an Ultramark 4 (Advanced Technology Laboratories) unit with a 7.5MHz mechanical sector transducer. In seven dogs, samples of renal cortex were obtained on 11 occasions using a percutaneous, ultrasound-guided biopsy technique [20]. In two other dogs, a wedge biopsy specimen of renal cortex was obtained during a celiotomy. Systemic blood pressures were determined in nine dogs on 13 occasions using a Dinamap Model 8300 (Critikon) oscillometric blood pressure monitor [35, 36]. For each blood pressure determination, five sequential Dinamap readings were averaged, and hypertension was present if systolic pressure exceeded 180 mm Hg or diastolic pressure exceeded 100 mm Hg. Hearing was tested 14 times in eight dogs by evaluating brainstem auditory-evoked responses (BAER) to high intensity (that is, 80 to 100 dB) clicks in each ear using a standard protocol [37]. The eyes of eight dogs were examined on 16 occasions by a veterinary ophthalmologist using a slit lamp biomicroscope and an indirect ophthalmoscope following dilation of the pupils. Urinalyses, complete blood counts, and serum biochemistry tests were performed using standard methods. Eighty urine specimens obtained from 17 dogs were evaluated. Urine protein:creatinine ratio (UPC) usually was determined, and for UPC, urine protein concentration was determined by a turbidometric method on a Discrete Clinical Analyzer (Dupont) and urine creatinine concentration was measured by the Ektachem Creatinine (singleslide method) two-point rate enzymatic test (Kodak). When three months of age, two of four dogs had slightly increased UPC values (up to 1.6) that did not persist as the dogs grew older [20]. Therefore, UPC $ 2 defined existence of proteinuria in this study. Hematuria was present if the Multistix (Ames) test result for blood was “small” or greater, or if as many as five erythrocytes were seen in a light microscopic field having a magnification of 3400. Association of hematuria with proteinuria was examined using a Fisher Exact Test (Statistix 4.1; Analytical Software, Tallahassee, FL, USA). Complete blood cell counts were performed on 47 specimens from 15 dogs. Platelet number was determined in 43 specimens from 13 dogs, but platelet size was not measured. A panel of serum chemistry tests was performed on 51 samples from 15 dogs.

Name

Type

Specificity

MAb M3F7 MAb 102 MAb A2 Anti-a4(IV) MAb A7 MAb H52 MAb B66 MAb C4

Monoclonal Monoclonal Monoclonal Polyclonal Monoclonal Monoclonal Monoclonal Monoclonal

a1/a2(IV) helix a1(IV) NC1 a3(IV) NC1 a4(IV) NC1 a5(IV) NC1 a5(IV) NC1 a6(IV) NC1 Laminin b2

MAb B1

Monoclonal

Laminin b1

MAb 2E8

Monoclonal

Laminin g1

Reference or source [38] [39] [40] [41] [42] [43] a

Hybridoma Bank, Iowa City, IA Chemicon, Temecula, CA Hybridoma Bank, Iowa City, IA

a

MAb B66 is a mouse monoclonal antibody raised against bovine a6(IV) NC1 by methods described in reference 44. This antibody crossreacts with human and dog a6(IV) chains.

Postmortem evaluations Nine of 10 affected dogs were euthanized and necropsied immediately thereafter at the Texas Veterinary Medical Center. One affected dog was euthanatized and necropsied in another state, and the dog’s kidneys were sent to Texas for evaluation. Histological and ultrastructural examination of the kidneys Histological examination. Kidney specimens obtained from 16 dogs were examined by light microscopy, and in three affected and one unaffected dogs, a series of renal specimens (2 to 4 from each dog) was examined. Tissues were fixed in 10% buffered formalin and embedded in paraffin. Sections approximately 3 mm thick were prepared with hematoxylin and eosin (H&E), periodic-acid schiff (PAS), Gomori’s methenamine silver (GMS) and trichrome stains for examination. Ultrastructural examination. Renal cortex from 16 dogs, including serial specimens from four dogs, were fixed in Karnovsky’s fixative, 4% paraformaldehyde and 6.25% glutaraldehyde in 0.1 M sodium cacodylate buffer, postfixed in 1% osmium tetroxide and embedded in EM BED 812 and Araldite 502 (Electron Microscopy Sciences, Fort Washington, PA, USA). Thin sections were cut on a Sovrall Ultra Microtome MT 2, supported on 300 mesh copper grids, and stained with saturated uranyl acetate and Sato’s lead citrate. Grids were examined in a Zeiss 10 C transmission electron microscope. Immunohistochemical procedures Tissues. Kidney specimens were obtained by biopsy or at necropsy from 12 dogs. Skin specimens also were obtained by biopsy or at necropsy from two affected and two normal dogs. Antibodies. The primary antibodies used in this study are listed in Table 2. All secondary antibodies were affinity-

709

Lees et al: Autosomal recessive Alport syndrome in dogs

purified fluorescein isothiocyanate-(FITC)-conjugated immunoglobulins, including goat anti-rabbit IgG (Biosource, Camarillo, CA, USA), goat anti-rat IgG (Pel-Freeze, Rogers, AR, USA), and goat anti-mouse IgG (Cappel, Durham, NC, USA). All secondary antibodies were absorbed with normal canine serum. Indirect immunofluorescence. Most tissues were embedded in Tissue-Tek OCT (Miles) and snap-frozen in liquid nitrogen, but two skin biopsy specimens were snap-frozen in isopentane that was precooled in liquid nitrogen. Tissues were sectioned at 4 mm in a Lipshaw cryostat in a constant temperature (25°C) and humidity (30%) room, and fixed for 10 minutes in acetone (for nondenatured sections) or ethanol (for sections to be denatured). After washing in fresh buffer (0.01 M PBS, pH 7.4), slides were incubated in a moist chamber with the appropriate dilution of primary antibodies. Incubation with FITC secondary antibodies was then performed. Slides that were to be incubated with anti-a4(IV), MAb A7 or MAb H52 were pretreated with 0.1 M glycine, 6 M urea, pH 3.5 for 10 minutes, then rinsed with distilled water and processed for the immunolabeling as described. To retard fluorescence quenching, a mounting media containing p-phenylenediamine was used. Labeling was examined with an epifluorescence microscope with appropriate filters (Carl Zeiss, Inc., Oberkochen, Germany), and fluorescence was graded 0 to 31 by two observers. Tissue sections directly incubated with secondary antibodies served as controls, and negative results were obtained in all instances. RESULTS Clinical features Affected dogs were 8 to 27 months (median, 13 months) of age when euthanized (Table 1; dogs 1 to 10). Two dogs were euthanized shortly after azotemia first occurred; the nephropathy in these dogs was not permitted to reach end stage. The other eight affected dogs were euthanized at or near the end stage of their disease. Early clinical features of HN were characterized in three affected dogs that were repeatedly evaluated beginning when they were three months old. At three to four months of age, the dogs appeared healthy. They exhibited good urine concentrating ability (urine specific gravity . 1.035), and proteinuria was absent (UPC , 2.0). Proteinuria, which began at five to eight months of age, was the first abnormality detected in affected dogs. A similar pattern of abnormalities was observed in affected dogs after they developed proteinuria, but the rate of subsequent change varied among the dogs. Growth rate decreased (data not shown), while the urine concentrating ability gradually diminished (data not shown), and concentrations of BUN and serum creatinine progressively increased (data not shown). Intervals from onset of proteinuria to development

Table 3. Hematuria in English cocker spaniel dogs with hereditary nephritis Number of specimens Proteinuric Non-proteinuric Total

Hematuric

Non-hematuric

Total

19 (24%) 5 (6%) 24 (30%)

21 (26%) 35 (44%) 56 (70%)

40 (50%) 40 (50%) 80 (100%)

of azotemia (serum creatinine . 2.0 mg/dl) were two to nine months. Two dogs that were later found to have HN exhibited proteinuria that was substantial but transient and associated with a brief episode of self-limiting, nonrenal illness. The proteinuria promptly resolved when the dogs recovered from their illnesses, and urine specimens (2 from each dog) that were evaluated during the next two months did not contain excess protein. Clinical signs were not observed in conjunction with the subsequent onset of persistent proteinuria in these dogs. Indeed, onset of persistent proteinuria was not associated with clinical signs in any affected dog. Excess protein was observed in 40 of 80 urine specimens, and every proteinuric sample was obtained from an affected dog. The 40 specimens that lacked proteinuria included all 30 samples obtained from seven dogs that did not have HN and 10 specimens obtained from four affected dogs before their onsets of persistent proteinuria. Values for UPC determined in 35 specimens obtained from 9 of 10 affected dogs with persistent proteinuria were 2.0 to 13.2 (median, 8.4). The highest UPC value observed in each affected dog was 7.0 to 13.2 (median, 11.1). Thirty-six specimens obtained from dogs that did not have HN (26 specimens), as well as from four affected dogs during periods before their onsets of proteinuria (10 specimens), had UPC values that were 0.05 to 1.6 (median, 0.31). Hematuria was observed in 5 (12.5%) of 40 specimens obtained from dogs that were not exhibiting proteinuria, and in 19 (47.5%) of 40 specimens from affected dogs with proteinuria. Hematuria was associated (P 5 0.001) with proteinuria (Table 3). Results of routine hematologic tests were within normal limits for specimens obtained from dogs that did not have renal failure (serum creatinine , 2.0 mg/dl). However, evaluations of affected dogs shortly before euthanasia invariably demonstrated anemia (hematocrit 11 to 33%; median, 20%; reference range, 37 to 55%). Three dogs that were severely uremic when they were evaluated had mild thrombocytopenia (125,000 to 192,000 platelets/ml; reference range, 200,000 to 500,000 platelets/ml), but platelet numbers were otherwise normal. Serum chemistry test results were abnormal only in affected dogs. Eight of 10 affected dogs had a reduced serum albumin concentration on at least one occasion. The

710

Lees et al: Autosomal recessive Alport syndrome in dogs

Fig. 1. (A) Light microscopic appearance of a kidney biopsy from a 6-month-old affected dog (no. 8) with proteinuria (Masson’s trichrome, 3600). Multiple foci of mesangial expansion are visible in the glomerulus (arrows). (B) Light microscopic appearance of a kidney biopsy from a 6-month-old affected dog (no. 9) with proteinuria (Masson’s trichrome, 3400). Multiple capillary walls within the glomerulus are thickened and have a “moth-eaten” appearance (arrows).

lowest serum albumin value observed in each dog was 1.8 to 3.0 mg/dl (median, 2.2 mg/dl; reference range, 2.4 to 3.6 mg/dl). Nine of 10 affected dogs had an increased serum cholesterol concentration on at least one occasion, and the highest cholesterol value observed in each dog was 218 to 395 mg/dl (median, 336 mg/dl; reference range, 120 to 247 mg/dl). Final evaluations of the affected dogs demonstrated varying degrees of azotemia (serum creatinine concentrations 2.1 to 12.8 mg/dl; median, 5.4 mg/dl; reference range, 0.3 to 2.0 mg/dl), increased serum phosphorus concentrations (7.6 to 25.2 mg/dl; median, 15.8 mg/dl; reference range, 2.9 to 6.2 mg/dl) and reduced serum total CO2 concentrations (5 to 20 mmol/liter; median, 16 mmol/liter; reference range, 21 to 28 mmol/liter). Blood pressure measurements were obtained from seven of ten affected dogs, one of three obligate carriers, and the unaffected dog. Hypertension was not detected in any dog. Abnormalities of kidney size, architecture or relative echogenicity were not detected by sonography in affected dogs that were not azotemic. Dogs with renal failure had hyperechoic renal cortices. Ultrasound-guided kidney biopsy procedures were performed without apparent complication. Signs suggestive of impaired hearing were not reported by the owners of any dog, and dogs appeared to respond normally to sound during their examinations. Hearing, as evaluated by BAER tests, was normal when evaluated in six of ten affected dogs, one of three obligate carriers, and the unaffected dog. Ocular abnormalities similar to lesions associated with Alport syndrome in humans were not observed in the eight dogs that were examined thoroughly. Postmortem examination of kidney Necropsies were performed on dogs with renal failure. Kidneys were uniformly reduced in size, overly firm, and

had fine pitting of their cortical surfaces. Renal capsules stripped easily. The kidneys were pale tannish-brown, the cortices were uniformly thinned, and the cortical to medullary ratios varied from 1:2.5 to 1:3 (normal 1:2). Histological examination of kidney Light microscopic lesions in affected dogs varied greatly depending on the stage of disease progression. The earliest lesion, which was observed in kidneys of six-month-old dogs, was focal mesangial expansion (Fig. 1A). Expanded mesangial areas stained weakly positive as collagen with Masson’s trichrome, but did not react with either the PAS or GMS techniques and was not visible through crossedpolars using Congo Red dye. Rare glomeruli were contracted with adhesions to Bowman’s capsule. Glomerular capillary basement membranes were segmentally thickened with both GMS and PAS, and the thickened areas had a “moth-eaten” appearance (Fig. 1B). The interstitium adjacent to affected glomeruli sometimes had infiltrates of lymphoid cells and mononuclear phagocytes. Five dogs that were necropsied at 9 to 17 months of age had multifocal cortical disease distributed in a radial pattern extending from the capsule to the corticomedullary junction with alternating zones of diseased and more normal parenchyma. In affected cortical tissue, glomeruli showed a range of lesions from mild to severe. Mild lesions consisted of segmental thickening of the capillary basement membranes and focal mesangial expansion similar to that described in six-month-old dogs. Moderate lesions were those of segmental glomerular fibrosis with adhesions to Bowman’s capsule and some early periglomerular fibrosis. Severe lesions consisted of complete glomerular sclerosis. In one dog, a majority of the remaining glomeruli were atrophic and contracted with cystic dilation of Bowman’s capsule.

Lees et al: Autosomal recessive Alport syndrome in dogs

711

Fig. 2. (A) Electron micrograph of glomerular basement membrane (GBM) from an unaffected 6-month-old dog (no. 13) showing normal thickness and ultrastructure (315,000). (B) Electron micrograph of GBM from an affected 6-month-old dog (no. 7) with proteinuria (315,000). A segment of the GBM is thinner than normal (arrows). (C) Electron micrograph of GBM from another area of the same specimen shown in panel B (315,000). The slightly thickened GBM exhibits bilaminar splitting and its epithelial aspect has an irregular contour. (D) Electron micrograph of GBM from an affected 15-month-old dog (no. 7) with renal failure showing marked thickening and multilamellation of the GBM (315,000).

The interstitium around diseased glomeruli was fibrotic and there were multifocal accumulations of lymphocytes and macrophages located around glomeruli and tubules. The latter displayed degenerative changes resulting in epithelial necrosis, tubular collapse, and fibrosis or tubular dilation with accumulation of proteinaceous material. Four dogs that were necropsied at 8 to 27 months of age had severe microscopic renal lesions similar to those described above, but the radial pattern of distribution was no longer visible. The lesions were homogeneously diffuse throughout the cortex, and there was mild to moderate medullary interstitial fibrosis. Ninety percent or more of glomeruli in any single section had lesions as described previously. However, more than 50% of glomeruli were sclerotic. Cystic glomerular atrophy was not prominent in these dogs. The magnitude of cellular inflammation was more pronounced in these dogs, as was the incidence of dilated tubules filled with proteinaceous material. There was also variable renal tubular, glomerular, and arterial intimal mineralization. The light microscopic appearance of kidney biopsy specimens from the remaining dogs was normal.

Ultrastructural examination kidney A range of ultrastructural changes were observed in the GBM of affected dogs examined at different stages of disease progression. In six-month-old dogs, samples obtained about one month after onset of persistent proteinuria had both thinning and thickening of the GBM (Fig. 2 A–C). Thickened GBM often exhibited bilaminar splitting and its epithelial aspect had an irregular contour. Focal fusion of visceral epithelial cell foot processes was associated with GBM thickening. These GBM abnormalities were diffuse but mild compared with later stages of progression. Subsequent specimens from these dogs, obtained about one month after onset of azotemia, exhibited multilamellation and greater thickening of the GBM with diffuse fusion of visceral epithelial cell foot processes. Especially in its thickest segments, the GBM was composed of a disorganized network of membranous strands that frequently enclosed electron-lucent areas. Spherical electron-dense structures were sometimes seen between the lamellations in the GBM. In dogs nearing the end stage of their disease, marked GBM thickening and multilamellation was extensive (Fig. 2D).

712

Lees et al: Autosomal recessive Alport syndrome in dogs Table 4. Relative distribution of a(IV) chains in renal and epidermal basement membranes of English cocker spaniel dogs with hereditary nephritis Distal tubular

Normal, HN-unaffected, and HN-carrier dogsa a1-a2(IV) a3-a4(IV) a5(IV) a6(IV) HN-affected dogsb a1-a2(IV) a3-a4(IV) a5(IV) a6(IV)

Epidermal

GBM

Bowman’s capsule

basement membrane

1 11 11 Variablec

11 2 1 1

11 11 11 1 (CT)

1 2 11 11

11 2 1 1

11 2 1 1

11 2 1 (CT) 1 (CT)

1 2 11 11

Abbreviations are: HN, hereditary nephritis; BM, basement membrane; CT, positive (1) labeling of BM in occasional tubules that appeared to be collecting tubules. a Dog nos. 12–16; N 5 5 b Dog nos. 3–5, 7–10; N 5 7 c Labeling of GBM was variable. In adolescent dogs, labeling was absent; however, positive (1) labeling was observed in dogs that were 30 months of age or older. Labeling was segmental in 2 dogs and diffuse in 2 others.

Fig. 3. Immunofluorescence staining for collagen IV chains in glomeruli of normal dogs (A, C, E, and G) and affected dogs (B, D, F, and H). (A and B) Comparison of labeling for a1(IV) chains in normal and affected dogs showing increased labeling of the thickened glomerular basement membrane (GBM) in the affected dog. (C and D) Comparison of labeling for a3(IV) chains in normal and affected dogs showing total absence of labeling in the affected dog. (E and F) Comparison of labeling for a4(IV) chains in normal and affected dogs showing total absence of labeling in the affected dog. (G and H) Comparison of labeling for a5(IV) chains in normal and affected dogs showing that labeling is present but reduced in the GBM of the affected dog. ™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3

In other dogs that were examined, GBM ultrastructure appeared normal. Immunohistochemical examinations Renal specimens. Similar patterns of immunofluorescence staining were observed in kidney from normal dogs, the unaffected dog, and obligate carriers (Table 4). The a1-a2(IV) chains were distributed in all basement membranes (BM) (Figs. 3A and 4A). Labeling for a3-a4(IV) chains was observed only in GBM and the BM of distal tubules (Figs. 3C, 3E, 4C, and 4E). The a5(IV) chain was co-distributed with a3-a4(IV) chains in GBM and distal tubular BM (Figs. 3G and 4G), but labeling for a5(IV) chains also was observed in BM of Bowman’s capsule (Fig. 3G) and collecting tubules and in the walls of renal arterioles (not shown). Variable labeling for a6(IV) chains was observed in GBM. Adolescent dogs had no GBM labeling for a6(IV) (Fig. 5A), but older dogs had segmental (Fig. 5B) or diffuse (Fig. 5C) GBM labeling for a6(IV) chains. In the unaffected dog, for example, GBM labeling for the a6(IV) chain was absent at 6 and 13 months of age but present at 30 and 50 months of age. The a6(IV) chain also was consistently co-distributed with the a5(IV) chain in BM of Bowman’s capsule (Figs. 5 A–C), in collecting tubular BM, and in the walls of renal arterioles (not shown). Labeling for each a(IV) chain in tubular BM generally was either strong or absent; however, the BM of

proximal tubules was weakly labeled for the a5(IV) chain by MAb A7, while labeling of proximal tubular BM for the a5(IV) chain by MAb H52 was totally absent. Labeling for laminin b2 was observed in GBM and Bowman’s capsule BM, but not in tubular BM. Labeling for laminin b1 was seen in BM of Bowman’s capsule and all tubules, but not in GBM. Labeling for laminin g1 was observed in all renal BM. The pattern of immunofluorescence staining observed in kidney from affected dogs was the same in every case, and labeling for each a(IV) chain was different than that observed in normal dogs (Table 4). Expression of the a1-a2(IV) chains in the GBM of affected dogs was increased (Fig. 3B), and labeling for the a1-a2(IV) chains was also observed in expanded interstitial spaces between tubules. Labeling for the a3-a4(IV) chains was totally absent both in the GBM and tubular BM (Figs. 3D, 3F, 4D, and 4F). Labeling for the a5(IV) chain was observed in GBM; however, intensity of fluorescence was reduced compared with that observed in normal dogs (Fig. 3H). Labeling for the a5(IV) chain was totally absent in the BM of most tubules, but strong labeling was observed in the BM of a few tubules that appeared to be collecting tubules (Fig. 4H). In affected dogs, labeling for a5(IV) chains in tubular BM by both MAb A7 and MAb H52 was always the same. The weak labeling of proximal tubular BM by Mab A7 seen in normal dogs was not observed in affected dogs. Labeling

Lees et al: Autosomal recessive Alport syndrome in dogs

713

714

Lees et al: Autosomal recessive Alport syndrome in dogs

Fig. 4. Immunofluorescence staining for collagen IV chains in renal tubules of normal dogs (A, C, E, and G) and affected dogs (B, D, F, and H). (A and B) Comparison of labeling for a1(IV) chains in tubular basement membrane (BM) of normal and affected dogs. In the affected dog, labeling also is observed in the expanded interstitial spaces between tubules. (C and D) Comparison of labeling for a3(IV) chains in tubular BM of normal and affected dogs showing total absence of labeling in the affected dog. (E and F) Comparison of labeling for a4(IV) chains in tubular BM of normal and affected dogs showing total absence of labeling in the affected dog. (G and H) Comparison of labeling for a5(IV) chains in tubular BM of normal and affected dogs showing labeling of the BM of occasional tubules that appear to be collecting tubules in the affected dog. ™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3

for the a5(IV) chain was observed in BM of Bowman’s capsule of affected dogs (Fig. 3H), as well as in the walls of their renal arterioles (not shown). Intensity of fluorescence labeling for the a5(IV) chain in these sites was similar to that observed in normal dogs. Additionally, GBM labeling for the a6(IV) chain was observed in all affected dogs (Fig. 5D). The a6(IV) chain was co-expressed with the a5(IV) chain in the BM of Bowman’s capsule (Fig. 5D), in collecting tubular BM (Fig. 5E), and in walls of renal arterioles (not shown). Except for observation of increased labeling for laminin g1 in expanded interstitial spaces between tubules, labeling for all the laminins examined was the same in affected dogs as was observed in normal dogs. Skin specimens. In normal dogs, labeling for a1-a2(IV) chains and for a5-a6(IV) chains was observed in epidermal BM, and labeling for a3-a4(IV) chains was totally absent (Table 4). In skin specimens from affected dogs, labeling for a1-a2(IV) chains (Fig. 5H) and for a5-a6(IV) chains (Fig. 5 F, G) was observed in the epidermal BM, with the intensity of fluorescence labeling the same as that seen in normal dogs. DISCUSSION We studied a progressive nephropathy that caused proteinuria and juvenile-onset chronic renal failure in 10 ECS dogs. The clinical and histological features of the kidney disease in these dogs were identical to those of the inherited nephropathy described previously in this breed [21– 26]. Some of the dogs we studied were siblings, and disease occurrence in their families was consistent with autosomal recessive transmission, which is the reported mode of inheritance for familial nephropathy in ECS dogs [27]. Pedigrees of all affected dogs showed that they had some ancestors in common, but this is probably true of many ECS dogs in North America, regardless of their health status. Because of this common ancestry, it is likely that the causative gene mutation is identical by descent in all ECS dogs with this inherited renal disease. Based on their ultrastructural GBM abnormalities and immunohistochemical features, the ECS dogs we studied were shown to have a form of HN, which we compared with other forms of HN reported in dogs and with AS in humans. Although hematuria was observed much more often in affected than in unaffected ECS dogs, hematuria was not as prominent a finding in the affected dogs as was proteinuria. Fewer than half of the specimens obtained from affected dogs showed hematuria, which was always mild, but all affected dogs exhibited persistent proteinuria,

which was substantial. Autosomal recessive HN in ECS dogs is thus similar to all previously reported forms of canine HN in that proteinuria rather than hematuria is the principal abnormality of the urine in affected dogs [15, 18, 45]. Additionally, certain transgenic COL4A3 mutant mice have proteinuria without hematuria [33], and humans with AS typically have persistent microscopic hematuria without proteinuria [1–3]. However, proteinuria eventually develops as the disease progresses, and correlation between the percent of GBM with lamina densa splitting and degree of proteinuria has been reported in patients with AS [46]. The clinicopathologic progression and severity of HN in affected male and female ECS dogs was similar, as expected for an autosomal recessive disease. Prior descriptions of familial nephropathy in ECS dogs state that the onset of illness usually occurs at 6 to 24 months of age, and the dogs we studied were euthanatized at 8 to 27 months of age. Thus, affected ECS dogs appear to survive somewhat longer than affected male Samoyed dogs with X-linked HN, which usually progress to terminal renal failure by 8 to 15 months of age [12, 15], but not so long as many bull terriers with autosomal dominant HN, which are reported to be a few months to 10 years old when death occurs [18]. In humans with AS, the rate of progression to end-stage renal disease also has significant interkindred variability, but X-linked Alport families are fairly cleanly divided into those in which the mean age of onset for end-stage renal disease in affected males is less or greater than 30 years [3]. Among dogs with HN, the rates of disease progression in male Samoyeds with X-linked disease and in ECS dogs with autosomal recessive HN are analogous to that observed in the subset of Alport patients who usually develop end-stage renal disease before the end of their third decade of life. Our evaluations of ECS dogs with autosomal recessive HN did not detect extrarenal abnormalities similar to those reported in humans with AS. The dogs did not have hearing deficits, ocular lesions or hematologic defects that seemed analogous to the entities that have been associated with the nephropathy in Alport patients. Such abnormalities generally have not been observed in dogs with other forms of HN; however, anterior lenticonus was reported in bull terriers with the autosomal dominant form [18]. The ultrastructural appearance of GBM in ECS dogs with autosomal recessive HN was identical to that seen in other forms of canine HN and in human AS. In studies of affected male Samoyed dogs with X-linked HN, electron microscopy demonstrated foci of bilaminar splitting of the GBM at one month of age, before the onset of proteinuria

Lees et al: Autosomal recessive Alport syndrome in dogs

715

716

Lees et al: Autosomal recessive Alport syndrome in dogs

Fig. 5. Immunofluorescence staining for collagen IV chains in glomeruli of normal dogs (A, B, and C) and in kidney (D and E) and skin (F, G, and H) of affected dogs. (A, B and C) Examples of the variable labeling for a6(IV) chains in glomerular basement membrane (GBM) observed in normal dogs. Adolescent dogs lack expression of the a6(IV) chain in their GBM (A), but older dogs have segmental (B) or diffuse linear (C) GBM labeling. Bowman’s capsule is labeled in all cases. (D) Labeling for a6(IV) chains in GBM and Bowman’s capsule of an affected dog. (E) Labeling for the a6(IV) chain in tubular BM showing labeling of occasional tubules that appear to be collecting tubules in an affected dog. (F) Labeling for the a6(IV) chain in epidermal BM of an affected dog. (G) Labeling for the a5(IV) chain in epidermal BM of an affected dog. (H) Labeling for the a1(IV) chain in epidermal BM of an affected dog. Intensity of epidermal BM labeling for collagen IV chains in affected dogs (panels F, G, and H) is the same as observed in normal dogs (not shown). ™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3

[13]. Subsequently, abnormal areas of GBM became multilaminar and confluent by three to four months of age, at which time proteinuria was persistent. In this study of ECS dogs with autosomal recessive HN, GBM ultrastructure was examined only in disease stages occurring after the development of persistent proteinuria. At the earliest stage examined, GBM alterations were diffuse and the predominant change was an irregular degree of bilaminar splitting and mild thickening, but some areas where the GBM appeared thinned were observed as well. Progressively greater degrees of thickening and multilamellation of the GBM were found at later stages of the disease. With the X-linked form of HN in Samoyed dogs, carrier females develop proteinuria at the same age as affected males [15, 16]. In such carrier females, electron microscopy reveals a mixture of GBM with multilaminar splitting and GBM with normal ultrastructure [13]. In contrast with the abnormalities exhibited by carrier female Samoyeds with X-linked HN, the clinical findings and ultrastructural appearance of the GBM in obligate ECS dog carriers of autosomal recessive HN were normal. Transgenic mice heterozygous for COL4A3 mutations also display no detectable abnormalities [33, 34]. Distribution of a(IV) chain expression observed in BMs of kidney from normal ECS dogs was similar to that found in normal humans [28, 39, 40, 47– 49], however, several differences were observed. One minor difference was that the a3-a4(IV) chains were not expressed in the Bowman’s capsule of ECS dogs, whereas these chains normally are expressed in portions of Bowman’s capsule (that is, near the vascular pole) of humans. Absence of the a3(IV) chain in dog Bowman’s capsule has been noted previously [18]. A more striking difference was that the a6(IV) chain was expressed in the GBM of normal adult ECS dogs. In normal human kidneys, the a6(IV) chain is not expressed in GBM [48, 49]. Additionally, the a6(IV) chain was not expressed in distal tubular BM of normal ECS dogs, whereas this chain normally is expressed in distal tubular BM of humans. Finally, expression of a5(IV) and a6(IV) chains was observed in the walls of renal arterioles of normal ECS dogs, but these chains are not expressed in the walls of renal arterioles of humans. The probe, B66, that we used for immunofluorescence staining of the a6(IV) chain is a specific anti-peptide antibody raised against bovine a6(IV) using previously described methods [44]. The distribution of renal B66 labeling was different than that observed for antibodies that

reacted with each of the other a(IV) chains. Compared with staining by antibodies for the a1-a2(IV) chains, a lack of B66 labeling was observed in GBM of adolescent normal dogs as well as in proximal and distal tubular BM of all normal and affected dogs. Compared with staining by antibodies for the a3-a4-a5(IV) chains, a lack of B66 labeling was observed in GBM of adolescent normal dogs and in distal tubular BM of all normal dogs. Additionally, B66 stained Bowman’s capsule and epidermal BM of all normal and affected dogs, but antibodies for a3-a4(IV) chains did not react with these structures in any dog. Complete absence of labeling for a3(IV) and a4(IV) chains in kidney of affected ECS dogs is consistent with the autosomal recessive mode of disease transmission in this breed, and is identical to findings in humans and mice with autosomal recessive AS [32–34]. The COL4A3 and COL4A4 genes presumably are located on an autosome in dogs as they are in humans and all other species studied to date. The finding of labeling for a5(IV) chains in affected ECS dogs at sites where a5(IV) chains are not normally co-expressed with a3(IV) and a4(IV) chains (that is, in Bowman’s capsule and epidermal BM), together with observation of similar disease frequency and severity in males and females, suggests that the COL4A5 gene is not mutated in these dogs. From the outset of our investigations of autosomal recessive HN in ECS dogs, we have theorized that the causative gene mutation would involve either the COL4A3 or the COL4A4 locus [19], and results of this study further support that hypothesis. Diffuse linear labeling for a1-a2(IV) chains, the a5(IV) chain, and the a6(IV) chain was observed in the thickened GBM of every affected dog examined. Presence of the a5(IV) chain in GBM of affected dogs was confirmed with two specific antibodies. Compared with the intensity of staining observed in the GBM of normal and obligate carrier ECS dogs, relative expression of the a5(IV) chain in GBM of affected dogs was markedly decreased and that of a1-a2(IV) chains and the a6(IV) chain was increased. A similar increase in expression of a1-a2(IV) chains has been observed previously in GBM of humans with X-linked AS [30], autosomal recessive AS [32] and COL4A3 mutant mice [33]. In humans with autosomal recessive AS, however, expression of the a5(IV) chain has not been observed in GBM that lacked expression of a3-a4(IV) chains. Our findings suggest that in dog GBM a5(IV) chains can be expressed in the absence of a3(IV) and a4(IV) chains. Glomerular cells that produce GBM also appear to be

Lees et al: Autosomal recessive Alport syndrome in dogs

717

718

Lees et al: Autosomal recessive Alport syndrome in dogs

capable of expressing a6(IV) chains as long as the a5(IV) chains are present, at least in ECS dogs. Our observations also suggest that an absence of a3(IV) and a4(IV) chains is sufficient to trigger the processes that cause multilamellation of the GBM to develop in dogs with HN. Diagnosis of autosomal recessive HN in ECS dogs can be supported by demonstration of absent a3-a4(IV) chain expression in kidney specimens; however, affected ECS dogs cannot be identified by immunohistochemical evaluation of their skin because a3-a4(IV) chains are not normally expressed in epidermal BM of dogs. Additionally, a(IV) chain expression in the kidney of obligate ECS dog carriers of HN is similar to that observed in normal dogs. Therefore, immunohistochemical evaluations are not useful for identification of HN-carriers in the ECS dog population. In Samoyed dogs with X-linked HN, evidence of decreased expression of the a3-a4-a5(IV) chains in the kidney of affected males has been demonstrated at the message and protein levels [50]. Among reported autosomal models of AS in dogs, however, HN in ECS dogs is the first disorder in which abnormal renal expression of a3a5(IV) chains has been demonstrated. In bull terriers with autosomal dominant HN, expression of a3(IV) and a5(IV) chains appeared to be normal by immunohistochemical evaluation [18]. Animal models of autosomal recessive AS have been generated in mutant mice [33, 34] but ECS dogs with HN are the only naturally occurring animal model of autosomal recessive AS yet described. Further investigation of this model may provide insights regarding the pathogenesis of ultrastructural GBM lesions in AS, because affected ECS dogs have typical thickening and multilamellation of GBM that contains a1(IV), a2(IV), a5(IV) and a6(IV) chains.

5.

6.

7.

8.

9. 10.

11. 12.

13.

14.

15.

ACKNOWLEDGMENTS This study was supported by grants from the American Veterinary Medical Foundation, the American Animal Hospital Association Foundation, the American Kennel Club Canine Health Foundation, and the National Institutes of Health. Anti-a4(IV) antibody was generously provided by J. Miner and J.R. Sanes, St. Louis, MO. The authors are grateful for superb technical assistance provided by Mr. Miles Frey and Ms. Kathy Divine.

16.

17.

Reprint requests to George E. Lees, D.V.M., Small Animal Medicine and Surgery, College of Veterinary Medicine, Texas A&M University, College Station, Texas 77843-4474, USA.

18.

REFERENCES

19.

1. KASHTAN CE, MICHAEL AF: Perspectives in clinical nephrology: Alport syndrome. Kidney Int 50:1445–1463, 1996 2. KASHTAN CE, SIBLEY RK, MICHAEL AF, VERNIER RL: Hereditary nephritis: Alport syndrome and thin glomerular basement membrane disease (chapt 39), in Renal Pathology (2nd ed), edited by TISHER CC, BRENNER BM, Philadelphia, J.B. Lippincott, 1994, pp 1239 –1267 3. GREGORY M, ATKIN CL: Alport syndrome (chapt 19), in Diseases of the Kidney, edited by SCHRIER RW, GOTTSCHALK CW, Boston, Little, Brown, 1993, pp 571–591 4. TRYGGVASON K, HEISKARI N: Alport syndrome (chapt 53), in Molecular Nephrology: Kidney Function in Health and Disease, edited by

20. 21. 22. 23.

SCHLONDORFF D, BONVENTRE JV, New York, Marcel Dekker, Inc., 1995, pp 795– 808 MOCHIZUKI T, LEMMINK HH, MARIYAMA M, ANTIGNAC C, GUBLER MC, PIRSON Y, VERELLEN-DUMOULIN C, CHAN B, SCHRODER CH, SMEETS HJ, REEDERS ST: Identification of mutations in the alpha3(IV) and alpha-4(IV) collagen genes in autosomal recessive Alport syndrome. Nature Genetics 8:77– 82, 1994 DING J, STITZEL J, BERRY P, HAWKINS E, KASHTAN CE: Autosomal recessive Alport syndrome: Mutation in the COL4A3 gene in a woman with Alport syndrome and posttransplant antiglomerular basement membrane nephritis. J Am Soc Nephrol 5:1714 –1717, 1996 LEMMINK HH, MOCHIZUKI T, VAN DEN HEUVEL LPWJ, SCHRODER CH, BARRIENTOS A, MONNENS LAH, VAN OOST BA, BRUNNER HG, REEDERS ST, SMEETS HJ: Mutations in the type IV collagen a3 (COL4A3) gene in autosomal recessive Alport syndrome. Hum Mol Genet 3:1269 –1273, 1994 KNEBELMANN B, FORESTIER L, DROUOT L, QUINOWES S, CHUET C, BENESSY F, SAUS J, ANTIGNAC C: Splice-mediated insertion of an Alu sequence in the COL4A3 mRNA causing autosomal recessive Alport syndrome. Hum Mol Genet 4:675– 679, 1995 FEINGOLD J, BOIS E, CHOMPRET A, BROYER M, GUBLER MC, GRU ¨ NFELD JP: Genetic heterogeneity of Alport syndrome. Kidney Int 27:672– 677, 1985 JEFFERSON JA, LEMMINK HH, HUNTER DD, HILL CM, SMEETS HJ, DOHERTY CC, MAXWELL AP: Autosomal dominant Alport syndrome linked to the type IV collagen alpha 3 and alpha 4 genes (COL4A3 and COL4A4). Nephrol Dial Transplant 12:1595–1599, 1997 JANSEN B, THORNER PS, SINGH A, PATTERSON JM, LUMSDEN JH, VALLI VE, BAUMAL R, BASRUR PK: Hereditary nephritis in Samoyed dogs. Am J Pathol 116:175–178, 1984 JANSEN B, TRYPHONAS L, WONG J, THORNER P, MAXIE MG, VALLI VE, BAUMAL R, BASRUR PK: Mode of inheritance of Samoyed hereditary glomerulopathy: An animal model for hereditary nephritis in humans. J Lab Clin Med 107:551–555, 1986 JANSEN B, THORNER P, BAUMAL R, VALLI V, MAXIE MG, SINGH A: Samoyed hereditary glomerulopathy (SHG): Evolution of splitting of glomerular capillary basement membranes. Am J Pathol 125:536 –545, 1986 THORNER P, JANSEN B, BAUMAL R, VALLI VE, GOLDBERGER A: Samoyed hereditary glomerulopathy: Immunohistochemical staining of basement membranes of kidney for laminin, collagen type IV, fibronectin, and Goodpasture antigen, and correlation with electron microscopy of glomerular capillary basement membranes. Lab Invest 56:435– 443, 1987 JANSEN B, VALLI VEO, THORNER P, BAUMAL R, LUMSDEN JH: Samoyed hereditary glomerulopathy: Serial clinical and laboratory (urine, serum biochemistry, and hematology) studies. Can J Vet Res 51:387–393, 1987 BAUMAL R, THORNER P, VALLI VEO, MCINNES R, MARRANO P, JACOBS R, BINNINGTON A, BLOEDOW AG: Renal disease in carrier female dogs with X-linked hereditary nephritis. Am J Pathol 139:751– 764, 1991 ZHENG K, THORNER PS, MARRANO P, BAUMAL R, MCINNES RR: Canine X chromosome-linked hereditary nephritis: A genetic model for human X-linked hereditary nephritis resulting from a single base mutation in the gene encoding the a5 chain of collagen type IV. Proc Natl Acad Sci USA 91:3989 –3993, 1994 HOOD JC, SAVIGE J, HENDTLASS A, KLEPPEL MM, HUXTABLE CR, ROBINSON WF: Bull terrier hereditary nephritis: A model for autosomal dominant Alport syndrome. Kidney Int 47:758 –765, 1995 LEES GE, WILSON PD, HELMAN GH, HOMCO LD, FREY MS: Glomerular ultrastructural findings similar to hereditary nephritis in 4 English cocker spaniels. J Vet Int Med 11:80 – 85, 1997 LEES GE, HELMAN GH, HOMCO LD, MILLICHAMP NJ, HUNTER JF, FREY M: Early diagnosis of familial nephropathy in English cocker spaniels. J Am Anim Hosp Assoc 34:189 –195, 1998 KROOK L: The pathology of renal cortical hypoplasia in the dog. Nord Vet-Med 9:161–176, 1957 PERSSON F, PERSSON S, ASHEIM A: Renal cortical hypoplasia in dogs: A clinical study on uraemia and secondary hyperparathyroidism. Acta Vet Scand 2:68 – 84, 1961 FREUDIGER U: Die kongenitale nierenrindenhypoplasie beim bunten cocker-spaniel. Schweizer Arch Tierheilk 107:547–566, 1965

Lees et al: Autosomal recessive Alport syndrome in dogs 24. STEWARD AP, MACDOUGALL DF: Familial nephropathy in the cocker spaniel. J Small Anim Pract 25:15–24, 1984 25. ROBINSON WF, HUXTABLE CR, GOODING JP: Familial nephropathy in cocker spaniels. Aust Vet J 62:109 –112, 1985 26. KOEMAN JP, EZILIUS JW, BIEWENGA WJ, BROM WE, GRUYS E: Zur familiaren nephropathie der cockerspaniel. Dtsch Tierarzti Wschr 96:174 –179, 1989 27. MACDOUGALL DF, NASH AS, CATTANACH BM: Control scheme for familial nephropathy in cocker spaniels. (letter) Vet Rec 121:134, 1987 28. KLEPPEL MM, KASHTAN CE, SANTI PA, WIESLANDER J, MICHAEL AF: Distribution of familial nephritis antigen in normal tissue and renal basement membranes of patients with homozygous and heterozygous Alport familial nephritis: Relationship of familial nephritis and Goodpasture antigens to novel collagen chains and type IV collagen. Lab Invest 61:278 –289, 1989 29. NAKANISHI K, YOSHIKAWA N, IIJIMA K, KITAGAWA K, NAKAMURA H, ITO H, YOSHIOKA K, KAGAWA M, SADO Y: Immunohistochemical study of a1-5 chains of type IV collagen in hereditary nephritis. Kidney Int 46:1413–1421, 1994 30. KASHTAN CE, KIM Y: Distribution of the a1 and a2 chains of collagen IV and of collagens V and VI in Alport syndrome. Kidney Int 42:115–126, 1992 31. YOSHIOKA K, HINO S, TAKEMURA T, MAKI S, WIESLANDER J, TAKEKOSHI Y, MAKINO H, KAGAWA M, SADO Y, KASHTAN CE: Type IV collagen a5 chain: Normal distribution and abnormalities in X-linked Alport syndrome revealed by monoclonal antibody. Am J Pathol 144:986 –996, 1994 32. GUBLER MC, KNEBELMANN B, BEZIAU A, BROYER M, PIRSON Y, HADDUM F, KLEPPEL MM, ANTIGNAC C: Autosomal recessive Alport syndrome: Immunohistochemical study of type IV collagen chain distribution. Kidney Int 47:1142–1147, 1996 33. MINER JH, SANES JR: Molecular and functional defects in kidneys of mice lacking collagen a3(IV): Implications for Alport syndrome. J Cell Biol 135:1403–1413, 1996 34. COSGROVE D, MEEHAN DT, GRUNKEMEYER JA, KORNAK JM, SAYERS R, HUNTER WJ, SAMUELSON GC: Collagen COL4A3 knockout: A mouse model for autosomal Alport syndrome. Genes Dev 10:2981– 2992, 1996 35. COULTER DB, KEITH JC: Blood pressures obtained by indirect measurement in conscious dogs. J Am Vet Med Assoc 184:1375–1378, 1984 36. MEURS KM, MILLER MW, SLATER MR: Comparison of the indirect oscillometric and direct arterial methods for blood pressure measurement in anesthetized dogs. J Am Anim Hosp Assoc 32:471– 475, 1996 37. BODENHAMER RD, HUNTER JF, LUTTGEN PJ: Brain stem auditoryevoked responses in the dog. Am J Vet Res 46:1787–1792, 1985 38. FOELLMER HG, MADRI JA, FURTHMAYR H: Monoclonal antibodies to type IV collagen: Probes for the study of structure and function of basement membranes. Lab Invest 48:639 – 649, 1983

719

39. BUTKOWSKI RJ, WIESLANDER J, KLEPPEL MM, MICHAEL AF, FISH AJ: Basement membrane collagen in the kidney: Regional localization of novel chains related to collagen IV. Kidney Int 35:1195–1202, 1989 40. KLEPPEL MM, MICHAEL AF: Expression of novel basement membrane components in the developing human kidney and eye. Am J Anat 187:165–174, 1990 41. MINER JH, SANES JR: Collagen IV a3, a4, and a5 chains in rodent basal laminae: Sequence, distribution, association with laminins, and developmental switches. J Cell Biol 127:879 – 891, 1994 42. DING J, KASHTAN CE, FAN W, KLEPPEL MM, SUN MJ, KALLURI R, NEILSON EG, MICHAEL AF: A monoclonal antibody marker for Alport syndrome identifies the Alport antigen as the a5 chain of type IV collagen. Kidney Int 46:1504 –1506, 1994 43. NAITO I, KAWAI S, NOMURA S, SADO Y, OSAWA G, JAPANESE ALPORT NETWORK: Relationship between COL4A5 gene mutation and distribution of type IV collagen in male X-linked Alport syndrome. Kidney Int 50:304 –311, 1996 44. SADO Y, KAGAWA M, KISHIRO Y, SUGIHARA K, NAITO I, SEYER J, SUGIMOTO M, OOHASHI T, NINOMIYA Y: Establishment by the rat lymph node method of epitope-defined monoclonal antibodies recognizing the six different a chains of human type IV collagen. Histochem Cell Biol 104:267–275, 1995 45. HOOD JC, ROBINSON WF, CLARK WT, SUTHERLAND RJ, JAMES I, THOMAS MAB, HUXTABLE CR: Proteinuria as an indicator of early renal disease in Bull terriers with hereditary nephritis. J Small Anim Pract 32:241–248, 1991 46. RUMPELT HJ: Hereditary nephropathy (Alport syndrome): Correlation of clinical data with glomerular basement membrane alterations. Clin Nephrol 13:203–207, 1980 47. KLEPPEL MM, SANTI PA, CAMERON JD, WIESLANDER J, MICHAEL AF: Human tissue distribution of novel basement membrane collagen. Am J Pathol 134:813– 825, 1989 48. PEISSEL B, GENG L, KALLURI R, KASHTAN CE, RENNKE HG, GALLO G, YOSHIOKA K, SUN MJ, HUDSON BG, NEILSON EG, ZHOU J: Comparative distribution of the a1(IV), a5(IV), and a6(IV) collagen chains in normal human adult and fetal tissues and in kidneys from X-linked Alport syndrome patients. J Clin Invest 96:1948 –1957, 1995 49. NINOMIYA Y, KAGAWA M, IYAMA K, NAITO I, KISHIRO Y, SEYER J, SUGIMOTO M, OOHASHI T, SADO Y: Differential expression of two basement membrane collagen genes, COL4A6 and COL4A5, demonstrated by immunofluorescence staining using peptide-specific monoclonal antibodies. J Cell Biol 130:1219 –1229, 1995 50. THORNER PS, ZHENG K, KALLURI R, JACOBS R, HUDSON BG: Coordinate gene expression of the a3, a4, and a5 chains of collagen type IV; Evidence from a canine model of X-linked nephritis with a COL4A5 gene mutation. J Biol Chem 271:13821–13828, 1996